U.S. patent number 5,884,242 [Application Number 08/895,620] was granted by the patent office on 1999-03-16 for focus spot detection method and system.
This patent grant is currently assigned to Micron Technology, Inc.. Invention is credited to Gregory King, Michael McMahon, Daniel Meier.
United States Patent |
5,884,242 |
Meier , et al. |
March 16, 1999 |
Focus spot detection method and system
Abstract
A method and system for detecting focus spots. Data from a file
created during stepper operation is extracted to get field
coordinate position, leveling scheme, and tilt with respect to the
x- and y-axes, and wafer height with respect to the focal plane for
the multiple fields on the multiple wafers in a production batch. A
delta value is calculated for the x- and y-axes tilt data which
averages the tilt of each field with its surrounding fields. Delta
values are placed in a 3-dimensional data structure linking
neighboring fields and corresponding fields on subsequent wafers.
Focus spots are detected by the repeated presence of data spikes
over the sum of the arithmetic mean and some multiple of the
standard deviation of the delta values.
Inventors: |
Meier; Daniel (Boise, ID),
King; Gregory (Meridian, ID), McMahon; Michael (Boise,
ID) |
Assignee: |
Micron Technology, Inc. (Boise,
ID)
|
Family
ID: |
24530641 |
Appl.
No.: |
08/895,620 |
Filed: |
July 17, 1997 |
Related U.S. Patent Documents
|
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
Issue Date |
|
|
631306 |
Apr 11, 1996 |
5673208 |
|
|
|
Current U.S.
Class: |
702/179; 414/935;
382/141; 382/144; 355/77; 355/55; 355/53; 702/81; 356/152.1;
356/141.2; 414/936; 382/145; 700/95; 700/109; 700/121; 700/110;
324/762.05 |
Current CPC
Class: |
G01J
1/4257 (20130101); G03F 9/7092 (20130101); G03F
9/70 (20130101); G03F 7/70641 (20130101); Y10S
414/137 (20130101); Y10S 414/136 (20130101); Y10S
414/135 (20130101); Y10S 414/14 (20130101) |
Current International
Class: |
G03F
7/20 (20060101); G01J 1/42 (20060101); G03F
9/00 (20060101); G01N 021/01 () |
Field of
Search: |
;702/179,81
;364/468.01,468.16,468.17,468.28 ;355/53,55,77 ;324/765
;356/141.2,152.1,400 ;414/247,935,937,940 ;382/141,144,145 |
References Cited
[Referenced By]
U.S. Patent Documents
Primary Examiner: Hafiz; Tariq R.
Assistant Examiner: Dam; Tuan Q.
Attorney, Agent or Firm: Schwegman, Lundberg, Woessner &
Kluth, P.A.
Parent Case Text
This application is a continuation of U.S. patent application Ser.
No. 08/631,306, filed Apr. 11, 1996, now U.S. Pat. No. 5,673,208.
Claims
We claim:
1. A method for detecting focus spots, comprising the steps of:
receiving a first plurality of data for a plurality of fields in a
plurality of substrates processed in a photolithography system;
organizing the first plurality of data into a data structure in a
memory;
calculating a second plurality of data from the first plurality of
data in the data structure;
storing the second plurality of data into the data structure;
setting an all-wafers threshold for indicating how many potential
focus spots must be identified throughout the plurality of
substrates before flagging a detected focus spot;
setting a same-spot threshold indicating how many potential focus
spots must be identified among corresponding coordinate fields of
the plurality of substrates before flagging a detected focus
spot;
providing an accumulator for tracking a number of fields
potentially having a focus spot;
calculating a second threshold to indicate the presence of a
potential focus spot by the second plurality of data;
traversing the data structure, wherein the step of traversing
comprises the steps of:
incrementing the accumulator if one of the second plurality of data
surpasses the second threshold;
reporting the detection of a focus spot if the accumulator
surpasses the all-wafers threshold; and
reporting the detection of a focus spot if the accumulator
surpasses the same-spot threshold.
2. The method for detecting focus spots as recited in claim 1,
wherein the step of traversing the data structure further comprises
the steps of:
traversing through the data structure horizontally among
consecutive fields on each of the plurality of substrates; and
traversing through the data structure vertically among
corresponding fields on each of the subsequent substrates when one
of the second plurality of data surpasses the second threshold.
3. The method for detecting focus spots as recited in claim 1,
further comprising the step of adjusting the lower bounds of the
first plurality of data to make the detection of focus spots more
reliable, the step of adjusting comprising the steps of:
calculating an arithmetic mean for the first plurality of data;
calculating a standard deviation for the first plurality of data;
and
replacing each of the data in the first plurality of data having a
value less than the arithmetic mean with the sum of the arithmetic
mean and the standard deviation.
4. The method for detecting focus spots as recited in claim 1:
wherein the step of storing the second plurality of data into the
data structure comprises the step of storing the second plurality
of data into a second data structure; and
wherein the step of traversing the data structure comprises the
step of traversing the second data structure.
5. The method for detecting focus spots as recited in claim 1,
wherein the step of calculating a second plurality of data
comprises the step of calculating a delta value for each of the
first plurality of data, and wherein the delta value for any data
of the first plurality of data for a field surrounded by
neighboring fields of similar data is the arithmetic mean of the
differences between the field and the neighboring fields.
6. The method for detecting focus spots as recited in claim 1,
wherein the step of organizing the first plurality of data into a
data structure in a memory comprises the step of linking the first
plurality of data for a field to the first plurality of data for
neighboring fields as well as to corresponding fields on previous
and subsequent substrates.
7. The method for detecting focus spots as recited in claim 1,
wherein the step of calculating a second threshold comprises the
step of calculating a multiple of a standard deviation above an
arithmetic mean.
8. A focus spot detector, comprising:
a data receiver for obtaining a first plurality of data for a
plurality of fields in a plurality of substrates processed in a
photolithography system;
an organizer for ordering the first plurality of data into a data
structure in a memory;
a calculator for computing a second plurality of data from the
first plurality of data and for storing the second plurality of
data into the data structure;
a first threshold for indicating a number of potential focus spots
detected for flagging a detected focus spot on one of the plurality
of substrates;
an accumulator for tracking the number of fields potentially having
a focus spot;
a second threshold for indicating the presence of a potential focus
spot by the second plurality of data;
a traversal engine for checking the second plurality of data,
wherein the traversal engine performs the steps of:
incrementing the accumulator if one of the second plurality of data
surpasses the second threshold; and
reporting the detection of a focus spot if the accumulator
surpasses the first threshold.
9. The focus spot detector as recited in claim 8 wherein the first
threshold for indicating a number of potential focus further
comprises:
an all-wafers threshold for indicating how many potential focus
spots must be identified throughout the plurality of substrates
before flagging a detected focus spot;
a same-spot threshold indicating how many potential focus spots
must be identified among corresponding coordinate fields of the
plurality of substrates before flagging a detected focus spot.
10. The focus spot detector as recited in claim 8 wherein the
traversal engine further performs the steps of:
traversing through the data structure horizontally among
consecutive fields for each of the plurality of substrates; and
traversing through the data structure vertically among
corresponding fields for each of the subsequent substrates when one
of the second plurality of data surpasses the second threshold.
11. The focus spot detector as recited in claim 8, further
comprising an invalidator that removes the first plurality of data
from focus spot detection.
12. The focus spot detector as recited in claim 8, further
comprising an adjustor for elevating the lower bounds of the first
plurality of data to make the detection of focus spots more
reliable, the adjustor performing the steps of:
calculating an arithmetic mean for the first plurality of data;
calculating a standard deviation for the first plurality of data;
and
replacing each of the data in the first plurality of data having a
value less than the arithmetic mean with the standard deviation
above the arithmetic mean.
13. The focus spot detector as recited in claim 8, further
comprising a second data structure; and
wherein the calculator for computing the second plurality of data
stores the second plurality of data in the second data
structure.
14. The focus spot detector as recited in claim 8, further wherein
the calculator for computing the second plurality of data computes
a delta value for each of the first plurality of data, wherein the
delta value for any data of the first plurality of data for a field
surrounded by neighboring fields of similar data is the arithmetic
mean of the differences between the field and the neighboring
fields.
15. The focus spot detector as recited in claim 8, wherein the
organizer for organizing the first plurality of data into a data
structure in a memory links the first plurality of data for a field
to the first plurality of data for neighboring fields as well as
for corresponding fields on previous and subsequent substrates.
16. The focus spot detector as recited in claim 8, wherein the
second threshold is a multiple of the standard deviation above the
arithmetic mean.
17. A computer program product, comprising:
a computer usable medium having computer readable program code
means embodied therein for causing a detection of focus spots, the
computer readable program code means in said computer program
product comprising:
computer readable program code means for causing a computer to
receive a first plurality of data for a plurality of fields in a
plurality of substrates processed in a photolithography system;
computer readable program code means for causing a computer to set
a first threshold for flagging a detected focus spot on one of the
plurality of substrates;
computer readable program code means for causing a computer to
provide an accumulator for tracking the number of fields
potentially having a focus spot;
computer readable program code means for causing a computer to
organize the first plurality of data into a data structure in a
memory;
computer readable program code means for causing a computer to
calculate a second plurality of data from the first plurality of
data in the data structure;
computer readable program code means for causing a computer to
store the second plurality of data into the data structure;
computer readable program code means for causing a computer to
calculate a second threshold to indicate the presence of a
potential focus spot;
computer readable program code means for causing a computer to
traverse the data structure;
computer readable program code means for causing a computer to
increment the accumulator if one of the second plurality of data
from the data structure surpasses the second threshold; and
computer readable program code means for causing a computer to
report the detection of a focus spot if the accumulator surpasses
the first threshold.
18. The computer program product as recited in claim 17, wherein
the computer readable program code means for causing a computer to
set a first threshold further comprising:
computer readable program code means for setting an all-wafers
threshold for indicating how many potential focus spots must be
identified throughout the plurality of substrates before flagging a
detected focus spot; and
computer readable program code means for setting a same-spot
threshold indicating how many potential focus spots must be
identified among corresponding coordinate fields of the plurality
of substrates before flagging a detected focus spot.
19. The computer program product as recited in claim 17, wherein
the computer readable program code means for causing a computer to
traverse the data structure further comprising:
computer readable program code means for traversing through the
data structure horizontally among consecutive fields for each of
the plurality of substrates; and
computer readable program code means for traversing through the
data structure vertically among corresponding fields for each of
the subsequent substrates when one of the second plurality of data
surpasses the second threshold.
20. The computer program product as recited in claim 17, further
comprising:
computer readable program code means for causing a computer to
invalidate the first plurality of data to remove a plurality of
fields from focus spot detection.
21. The computer program product as recited in claim 17, further
comprising:
computer readable program code means for causing a computer to
store the second plurality of data into a second data structure;
and
computer readable program code means for causing a computer to
increment the accumulator if one of the second plurality of data
from the second data structure surpasses the second threshold.
22. The computer program product as recited in claim 17, further
comprising:
computer readable program code means for causing a computer to
calculate a second plurality of data by calculating a delta value
for all of the data of the first plurality of data; wherein the
delta value for a field surrounded by neighboring fields of similar
data is the arithmetic mean of the differences between the field
and the neighboring fields.
23. A method for detecting focus spots, comprising the steps
of:
processing a plurality of substrates with a photolithographic
stepper for preparing a plurality of fields for each of the
plurality of substrates;
measuring a tilt with respect to the x-axis, a tilt with respect to
the y-axis, and a substrate height for producing therefrom a
plurality of measurement data for each of the plurality of
fields;
organizing the plurality of measurement data into a data
structure;
calculating an arithmetic mean and a standard deviation for the
plurality of measurement data;
setting a flagging threshold based on the arithmetic mean and the
standard deviation;
setting an all-wafers threshold, which indicates how many potential
focus spots must be identified before flagging a detected focus
spot;
setting a same-spot threshold, which indicates how many potential
focus spots must be identified among corresponding coordinate
fields of the plurality of substrates before flagging a detected
focus spot;
inspecting the plurality of measurement data for incrementing an
accumulator when the data from the plurality of measurement data
surpasses the flagging threshold; and
reporting the detection of a focus spot if the accumulator
surpasses a detection threshold.
24. The method for detecting focus spots, as recited in claim 23,
wherein the step of inspecting the plurality of measurement data
further comprises the steps of:
traversing through the data structure horizontally among
consecutive fields for each of the plurality of substrates; and
traversing through the data structure vertically among
corresponding fields for each of the subsequent substrates when one
of the second plurality of data surpasses the second threshold.
25. The method for detecting focus spots, as recited in claim 23,
wherein the step of organizing the plurality of measurement data
into a data structure comprises the step of calculating a delta
value for each of the plurality of measurement data; wherein the
delta value for data in the plurality of measurement data for a
field surrounded by neighboring fields of similar data is the
arithmetic mean of the differences between the field and the
neighboring fields.
26. The method for detecting focus spots, as recited in claim 23,
further comprising the steps of:
calculating a preliminary arithmetic mean and a preliminary
standard deviation for the plurality of measurement data;
adjusting the lower bounds of the plurality of measurement data by
replacing the plurality of measurement data evaluating to less than
the preliminary arithmetic mean with the sum of the preliminary
arithmetic mean and the preliminary standard deviation.
27. A computer for detecting focus spots, comprising:
a processor;
at least one data file, wherein each data file contains a plurality
of first data values corresponding to characteristics of a
processed wafer;
a computer-readable medium; and
at least one application program, each application program executed
by the processor from the computer-readable medium,
wherein each program calculates a plurality of second data values
from each data file to correspond to a plurality of predefined
coordinate positions on the wafer,
wherein the second data values corresponding to each predefined
coordinate position are compared to a predetermined threshold
level, such that a focus spot is detected if the threshold is
exceeded in a predetermined number of coordinate positions.
28. The computer of claim 27, wherein the second data values
corresponding to each predefined coordinate position are compared
among consecutive fields on each of the processed wafers.
29. The computer of claim 27, wherein the characteristics of the
processed wafer are determined via a stepper machine.
30. The computer of claim 29, wherein the characteristics of a
processed wafer are height and tilt of the wafer with respect to
the x-axes and y-axis.
31. The computer of claim 30 further comprises a stepper
machine.
32. The computer of claim 27, wherein the computer-readable medium
comprises a memory.
33. A computer for detecting focus spots, comprising:
a processor;
at least one data file, wherein each data file contains a plurality
of first data values corresponding to characteristics of a
plurality of processed wafers;
a computer-readable medium; and
at least one application program, each application program executed
by the processor from the computer-readable medium,
wherein each program calculates a plurality of second data values
from each data file to correspond to a plurality of shared
predefined coordinate positions on one of the plurality of
processed wafers,
wherein the second data values corresponding to the same coordinate
position of each one of the plurality of wafers are compared to a
predetermined threshold level, such that a focus spot is detected
if the threshold is exceeded in a predetermined number of compared
coordinate positions.
34. The computer of claim 33, wherein the second data values
corresponding to each predefined coordinate position are compared
vertically among corresponding fields on each of the subsequent
processed wafers when one of the second plurality of data surpasses
the threshold.
35. The computer of claim 33, wherein the characteristics of a
processed wafer are determined via a stepper machine.
36. The computer of claim 35, wherein the characteristics of a
processed wafer are height and tilt of the wafer with respect to
the x-axes and y-axis.
37. The computer of claim 36 further comprises a stepper
machine.
38. The computer of claim 33, wherein the computer-readable medium
comprises a memory.
39. A computer for detecting focus spots, comprising:
a processor;
at least one data file, wherein each data file contains a plurality
of first data values corresponding to characteristics of a
plurality of processed wafers;
a computer-readable medium; and
at least one application program, each application program executed
by the processor from the computer-readable medium,
wherein each program calculates a plurality of second data values
from each data file to correspond to a plurality of shared
predefined coordinate positions on one of the plurality of
processed wafers,
wherein the second data values corresponding to each predefined
coordinate position are compared to a predetermined threshold
level, if a threshold is exceeded in a predetermined number of
coordinate positions on a wafer, then
the second data values corresponding to the same coordinate
position of each one of the plurality of wafers are compared to a
predetermined threshold level, such that a focus spot is detected
if the threshold is exceeded in a predetermined number of compared
coordinate positions.
40. The computer of claim 39, wherein the second data values
corresponding to each predefined coordinate position are compared
among consecutive fields on each of the wafers, and wherein the
second data values corresponding to the same coordinate position of
each of the plurality of wafers are compared vertically among
corresponding fields on each of the subsequent wafers when one of
the second plurality of data surpasses the threshold.
41. The computer of claim 40, wherein the characteristics of a
processed wafer are height and tilt of the wafer with respect to
the x-axes and y-axis.
42. The computer of claim 39, wherein the characteristics of a
processed wafer are determined via a stepper machine.
43. The computer of claim 42 further comprises a stepper
machine.
44. The computer of claim 39, wherein the computer-readable medium
comprises a memory.
45. A computer program product comprising a memory having computer
program logic recorded thereon for enabling a processor in a
computer system to link information, the computer program logic
comprising:
a receiving process for enabling the processor to receive a first
plurality of data for a plurality of fields in a plurality of
substrates processed in a photolithography system;
a organizing process for enabling the processor to organize the
first plurality of data into a data structure in a memory;
a calculating process for enabling the processor to calculate a
second plurality of data from the first plurality of data in the
data structure;
a storing process for enabling the processor to store the second
plurality of data into the data structure;
an entering process for enabling the processor to enter an
all-wafers threshold for indicating how many potential focus spots
must be identified throughout the plurality of substrates before
flagging a detected focus spot;
an entering process for enabling the processor to enter a same-spot
threshold indicating how many potential focus spots must be
identified among corresponding coordinate fields of the plurality
of substrates before flagging a detected focus spot;
an accumulating process for enabling the processor to accumulate
for tracking a number of fields potentially having a focus
spot;
a calculating process for enabling the processor to calculate a
second threshold to indicate the presence of a potential focus spot
by the second plurality of data;
a traversing process for enabling the processor to traverse the
data structure, wherein the step of traversing comprises:
a incrementing process for enabling the processor to increment the
accumulator if one of the second plurality of data surpasses the
second threshold;
a reporting process for enabling the processor to report the
detection of a focus spot if the accumulator surpasses the
all-wafers threshold; and
a reporting process for enabling the processor to report the
detection of a focus spot if the accumulator surpasses the
same-spot threshold.
46. The computer program product of claim 45, wherein the
traversing process further comprises:
a traversing process for enabling the processor to traverse through
the data structure horizontally among consecutive fields on each of
the plurality of substrates; and
a traversing process for enabling the processor to traverse through
the data structure vertically among corresponding fields on each of
the subsequent substrates when one of the second plurality of data
surpasses the second threshold.
Description
FIELD OF THE INVENTION
The present invention relates to the production of integrated
circuits by photolithography and in particular to the inspection of
the wafers manufactured by photolithographic means to detect the
presence of focus spots.
BACKGROUND
Photolithography is used to produce semiconductor integrated
circuits (ICs). Such a process uses photographic techniques to
impart patterns for the construction of ICs on wafers. Due to the
extremely small size of these intricate circuits, the manufacturing
process must be excessively monitored to ensure that contaminating
particles are not present. A particle may be inadvertently
introduced to the process, often times becoming sandwiched between
the wafer and its retaining vacuum platform. The intruding particle
causes a region of the wafer to be subject to improper focusing by
the photographing unit.
Improper focusing on such a wafer region causes what is known as a
focus spot. Such spots are recognizable by human inspectors and are
a major problem for chip makers. Focus spots can affect perhaps 4%
of all batches. There is a need in the art for an automated system
for focus spot detection. Such a system could inspect a batch of
wafers soon after their production, thus saving the time lag
experienced as human inspectors more slowly examine the wafers.
Such a system could save expense as well as time. Frequently, a
surface contaminant will remain on the vacuum platform and will
cause focus spots in several wafers at the same location. By
detecting the appearance of a focus spot soon after the batch has
been completed, engineers can be notified to clean the vacuum
platform before the next batch of wafers is introduced.
The detection method must also be accurate. False focus spot
detection is detrimental to the manufacturing process because such
false flagging of batches will cause added work for human
inspectors, and will cause unnecessary interruption of the
manufacturing process for machine inspection and cleaning.
SUMMARY OF THE INVENTION
A method for detecting focus spots is described. A data file
produced during the stepper operation of the manufacturing of
batches of semiconductor wafers containing multiple exposure fields
is utilized. Leveling schemes, field coordinate positions, wafer
tilt with respect to the x- and y-axes (Rx and Ry, respectively),
and wafer height with respect to the focal plane are extracted from
the data file. The data is lower-bound adjusted by adjusting Rx, Ry
and wafer height values which fall below the arithmetic mean of
batch data and by calculating a delta value for each data element.
The delta value consists of a type of average of the value with
that of four neighboring values. The delta values are placed into a
3-dimensional data structure for analysis.
Focus spots are reported when the traversal of the data structure
finds delta values above a flagging threshold (the sum of the
arithmetic mean and some multiple of the standard deviation) for a
specified number of times in the same coordinate position or for a
specified number of times in general.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a block diagram of the photolithography process.
FIG. 2 is a block diagram of a computer system that tests for focus
spots.
FIG. 3 is a diagram of one wafer from a wafer batch.
FIG. 4 is a graphical representation of how neighboring fields
determine a delta value for a field.
FIG. 5 is a diagram of a wafer, showing horizontal traversal.
FIG. 6 is a diagram of a wafer batch, showing vertical
traversal.
FIG. 7 is a flowchart of the method of detecting focus spots.
DETAILED DESCRIPTION
In the following Detailed Description, reference is made to the
accompanying drawings which form a part hereof and in which is
shown by way of illustration specific embodiments in which the
invention may be practiced. These embodiments are described in
sufficient detail to enable those skilled in the art to practice
and to use the invention, and it is to be understood that other
embodiments may be utilized and that changes may be made without
departing from the spirit and scope of the present invention. The
following Detailed Description is, therefore, not to be taken in a
limiting sense and the scope of the invention is defined by the
appended claims. In the figures, elements having the same number
perform essentially the same functions.
Focus spots can spoil wafers manufactured by photolithography. FIG.
1 is a block diagram of the photolithography process. FIG. 1 shows
a photolithographic system 100 comprising a stepper machine 105
capable of working wafers 110.1 through 110.n. Stepper machine
includes an exposure unit 107, vacuum platform 115, processing unit
120 and physical control devices 125. Exposure unit 107 of current
technology includes an illumination source and pattern mask.
Exposure unit 107 of a future technology may utilize electron beams
instead to pattern the wafer directly. Vacuum platform 115
accommodates wafer 110 during the photolithographic exposure.
Physical control devices 125 include actuators which adjust the
wafer's height and tilt about the x- and y-axes. Processing unit
120 is programmed to control the movement of physical control
devices 125. Processing unit 120 also retrieves readings from
physical control devices 125 and writes such data to data file 130.
Stepper machine 105 can be a PAS 5500/100 stepper available from
ASM Lithography of Velthoden, Holland. Processing unit 120 can be
implemented as a Sun Workstation available from Sun Microsystems of
Palo Alto, Calif. Processing unit 120 can use ASM Lithography's PAS
software to obtain the readings from physical control devices 125
and save such data to data file 130. Other stepper machines,
processing units and software can also be used.
FIG. 2 is a block diagram of a computer system 200 that tests for
focus spots after photolithographic system 100 processes wafers
110.1 through 110.n. Computer system 200 can be a separate
processing unit from photolithographic system 100 or it can be
integrated with photolithographic system 100. Computer system 200
includes input devices 205, screen display 210, printer 215,
storage 220 and processing unit 225. Processing unit 225 can be a
Sun workstation available from Sun Microsystems in Palo Alto,
Calif. Storage 220 includes operating system 230, focus spot
detection engine 235, raw data 3-D data structure 240, delta data
3-D data structure 245, as well as data file 130 created by stepper
machine 105.
Focus spot detection engine 235 reads data file 130 containing data
from physical control devices 125 as they processed wafers 110.1
through 110.n. Focus spot detection engine 235 constructs raw data
3-dimensional data structure from data extracted from data file
130. Delta data 3-dimensional data structure 245 is a second data
structure created by focus spot detection engine 235 through
smoothing of the data. Focus spot detection engine 235 analyzes
resulting delta data 3-D data structure 245 for the presence of
focus spots in wafers 110.1 through 110.n.
FIG. 3 is a diagram of one wafer 110.n from a wafer batch. A batch
is made up of several wafers. Each wafer is made up of fields 305.1
through 305.n. Each field is the amount of wafer surface exposed at
each step of the photolithographic process. As each field 305 is
ready to be developed, physical control devices 125 set the wafer
height (called Z) and tilt with respect to the x- and y-axes
(called Rx and Ry, respectively) in order to bring field 305 into
optimal focus.
This procedure of adjusting the wafer is, in one embodiment of
stepper machine 105, accomplished by three leveling schemes.
Simply, the "All-by-LS" scheme is used on fields, such as field
305.3, which are generally in the center of wafer 110.n. These are
complete fields that do not fall off the edge of wafer 110.n.
All-by-LS fields have unique Z, Rx, and Ry data values. The
"Z-by-LS" scheme is used for fields at the wafer edge, such as
field 305.2. Such fields generally have one side falling off the
edge or fields with two edges falling off the wafer's edge that are
adjacent to two All-by-LS fields. Z data is extracted from these
Z-by-LS fields, but Rx and Ry data are not. The "None-by-LS" scheme
is used generally on corner fields, such as field 305.1, where two
sides fall off the edge of the wafer and that aren't adjacent to
two All-by-LS fields. No unique Z, Rx, or Ry data is derived from
None-by-LS fields.
Regardless of the scheme used, the Z, Rx and Ry data is part of the
total data stored by processing unit 120 to data file 130 for each
field 305 of each wafer 110 in a wafer batch. Focus spot detection
engine 235 uses Rx and Ry data to determine whether a focus spot is
present. Therefore, only All-by-LS fields, which have corresponding
Rx and Ry data are analyzed. Rather than using the raw data stored
in data file 130, focus spot detection engine 235 achieves better
results by calculating a delta value for each field's Rx and Ry
data. FIG. 4 is a graphical representation of how neighboring
fields determine a delta value for field 305. FIG. 4 shows five
fields 305.4 through 305.8. To determine the delta value X' for
field X 305.4, values from field A 305.5, field B 305.6, field C
305.7 and field D 305.8 are used to find the arithmetic mean of the
differences of the fields A, B, C and D from the field X. The
formula for a delta value X' is thus: ##EQU1##
In this manner a delta Rx and a delta Ry is calculated for each raw
Rx and raw Ry. This data smoothing is necessary to detect real
focus spots while avoiding false ones. Having smoothed data allows
focus spots to be more readily observed. The defining equation
indicating a possible focus spot is given as:
where .mu. is the arithmetic mean and .sigma. is the standard
deviation of all fields 305 in the wafer batch and x is some chosen
constant multiplier. Because focus spots occur when Rx and Ry
values spike, focus spots only occur when these values are above
the mean. To reduce data variability, raw data values below the
mean are adjusted to the value of (.mu.+.sigma.) prior to being
used to calculate the field deltas.
Focus spot detection engine 235 inspects the field deltas from
fields 305 in order to determine whether a focus spot is likely to
have occurred. This determination is accomplished by a dual
traversing algorithm. First, focus spot detection engine 235
inspects fields 305 horizontally. FIG. 5 is a diagram of a wafer,
showing horizontal traversal. In FIG. 5, each field 305 is
inspected sequentially. This is accomplished by the use of delta
data 3-D data structure 245 which is constructed so that the
sequential inspection shown in FIG. 5 is easily done. If any field
during this horizontal traversal has a delta Rx or delta Ry
surpassing the threshold limit (as defined by (.mu.+x.sigma.)),
then a possible focus spot has been detected. In order to lessen
the chance of a false focus spot report, a second traversing
algorithm is executed.
FIG. 6 is a diagram of a wafer batch, showing vertical traversal,
the second traversing algorithm. As noted above, frequently a
contaminant that causes a focus spot will remain on vacuum platform
115 during successive wafer 110 manufacturing. Therefore, a
potential focus spot may be better validated if a possible focus
spot is found in the same coordinate position of subsequent wafers.
In FIG. 6, the corresponding fields on successive wafers 110.1
through 110.4 are checked. This is called vertical traversing and
is assisted by the 3-dimensionality of delta data 3-D data
structure 245. If enough possible focus spots are found in the same
location on successive wafers during the vertical traversal, a
focus spot is reported. However, since sometimes a focus spot does
not appear in the same location on subsequent wafers, focus spot
detection engine 235 also reports focus spots if a threshold of
such singular observations is surpassed. Upon finding a focus spot
with a wafer batch, focus spot detection system 200 will display a
warning message on screen display 210.
FIG. 7 is a flowchart of the method of detecting focus spots
carried out by focus spot detection engine 235. In one embodiment,
the flowchart of FIG. 7 is written as a C program. Other computer
languages could be used to implement the flowchart as well. In FIG.
7, at step 703, the raw data is extracted from data file 130. This
raw data includes Rx, Ry, Z-height, leveling scheme, and coordinate
position for each field 305 of each wafer 110 in the wafer batch.
At step 706, the arithmetic mean (.mu.) and standard deviation
(.sigma.) are calculated for both Rx and Ry. Then the raw data is
organized into a 3-dimensional data structure at step 709. The data
structure links neighboring fields 305 and corresponding fields in
the previous and next wafers 110. Because focus spots occur when Rx
and Ry values spike, to smooth the data all Rx and Ry values in raw
data 3-D data structure 240 which are below .mu. are replaced with
(.mu.+.sigma.) at step 712.
After the low values are replaced, step 715 marks certain fields as
invalid for consideration. Such fields which are marked may be
those by which Z-by-LS and None-by-LS schemes were used to obtain
data from physical control devices 125 since these two schemes do
not generate unique Rx and Ry values. In other embodiments, and as
manufacturing techniques are improved, fewer or no fields may be
marked as invalid. The next step, step 718, is to create a second
data structure. The delta data 3-D data structure 245 is filled
with field delta values (X') as explained previously. Then at step
721, the arithmetic mean (.mu.) and standard deviation (.sigma.)
are calculated for the field delta.
Now that delta data 3-D data structure 245 exists, at step 724, the
flagging threshold (.mu.+x.sigma.) is determined. The SIMGALEVEL
parameter is the constant multiplier (x) to use in the flagging
threshold equation. This is set by the user. Often, it will be
between two and eight. The SIGMALEVEL is set with the issues of
increased accuracy and mitigating false flags in mind. A lower
SIGMALEVEL will cause more focus spots to be reported. However,
more of these reports will be false flags than if a higher
SIGMALEVEL is used.
With all data calculated, the traversals to seek a focus spot
begin. Primarily, wafers 110.1 through 110.n are checked by a
horizontal traversal, starting at step 727. The horizontal
traversal checks each field 305.1 through 305.n on each wafer 110.1
through 110.n sequentially. As field 305 is checked in step 730, it
is marked in step 733 so that field 305 will only be checked once.
If field 305's field delta is greater than the flagging threshold
then a vertical search is commenced at step 742.
A vertical traversal checks the field data for corresponding fields
at the same coordinate position in subsequent wafers 110. Similar
to the horizontal traversal, at step 745 the field data is checked
and at step 748 the field is marked so that it will only be checked
once. If the field delta value is greater than the flagging
threshold, at step 754 the number of times that the fields in the
coordinate position have exceeded the threshold are compared
against the SAMESPOT parameter. Once SAMESPOT is exceeded, the
system should report at step 766 that a focus spot was found. If
the SAMESPOT parameter is not exceeded, a fields.sub.-- flagged
counter is incremented at step 757. This counter will cause a focus
spot to be reported, at step 763, if the counter is greater than
the ALLWAFERS parameter.
The SAMESPOT parameter will usually be a small value, perhaps five.
This indicates that a potential focus spot located on the same
field in five subsequent wafers 110 will cause the system to report
a focus spot. The ALLWAFERS parameter will usually be larger than
SAMESPOT, perhaps 15. This parameter indicates that when 15
potential focus spots are found by the system, a focus spot will be
reported, even without corresponding focus spots on subsequent
wafers.
Other embodiments of the present invention are possible without
departing from the scope and spirit of the present invention. Other
embodiments of this invention include a configuration allowing the
currently invalidated fields on the edge of the wafers to be
included in the focus spot detection method.
Although specific embodiments have been illustrated and described
herein, it will be appreciated by those of ordinary skill in the
art that any arrangement which is calculated to achieve the same
purpose may be substituted for the specific embodiment shown. This
application is intended to cover any adaptations or variations of
the present invention. Therefore, it is manifestly intended that
this invention be limited only by the claims and the equivalents
thereof.
* * * * *